Bayesian Predictive Model for Electric Level 4 Connected Automated Vehicle Adoption

The electric Level 4 connected automated vehicles (CAVs) are now allowed in several cities around the world to demonstrate their automation capability in shared mobility robotaxi and microtransit services in geofenced areas. Private and public sector stake-holders need predictions of their adoption without regulatory constraints for personal mobility and use in shared mobility services. In anticipation of the future presence of CAVs in transportation vehicle fleets, governments are planning necessary regulatory and infrastructure changes. Accompanied with the need for forecasts is the acknowledgement that CAV adoption decisions must be made under uncertain states of technology and infrastructure readiness. This paper presents a Bayesian predictive modelling framework for electric Level 4 CAV adoption in the 2030-2035 application context. The inputs to the Bayesian model are obtained from effectiveness estimates of CAV applications that are processed with the Montecarlo method to accounts for uncertainties in these estimates. Scenarios of CAV adoption in the 2030-2035 period are analyzed using the Bayesian model, including the quantification of the value of new information obtainable from demonstration studies intended to reduce uncertainties in technology and infrastructure readiness. The results show that in the 2030-2035 application context, the CAVs are likely to be adopted, provided that the trajectory of progress in technology and infrastructure readiness continues, and potential adopters are offered opportunities to learn about Level 4 CAV technological capabilities in real life service environment. The findings of this research can be used by private and public sector interest groups.